The appressorium at a glance

To gain entry to host tissue, many fungal pathogens develop infection cells to breach the tough outer surface of a plant or animal (Mendgen et al., 1996; Talbot, 2019). These include hyphopodia, infection cushions and appressoria (Bozkurt and Kamoun, 2020; Choquer et al., 2021; Goos and Gessner, 1975; Ryder and Talbot, 2015; Talbot, 2019). Here, we focus on appressoria, which are the most-studied fungal infection cells (Chethana et al., 2021a,b), and are vital to many of the most destructive plant diseases (Dean et al., 2012).

The name appressorium was first coined in the 19th century by Albert Bernhard Frank who defined them as “a swelling on a germ tube or hypha, especially for attachment in an early stage of infection” (Frank, 1883). However, the function of appressoria was later found to be not only for attachment, but also to rupture the host cuticle (Deising et al., 2000). Appressoria are classified into two main groups, single-celled appressoria – including proto-appressoria, hyaline appressoria and melanised appressoria – and compound appressoria, such as infection cushions (see poster) (Chethana et al., 2021a). In addition to plant pathogens, appressoria can be formed by endophytes, saprobes and symbionts, including some arbuscular mycorrhizal fungi, as well as lichens and insect pathogenic fungi (Deising et al., 2000; Demoor et al., 2019; Hyde et al., 2020; Lozano-Tovar et al., 2013; Ryder and Talbot, 2015). The devastating rice blast fungus Magnaporthe oryzae (Eseola et al., 2021) elaborates appressoria (see poster), which melanise and generate enormous intracellular turgor to facilitate mechanical penetration of the rice leaf (Hyde et al., 2020; Mendgen et al., 1996; Ryder and Talbot, 2015). Non-melanised (or slightly melanised) appressoria are formed by other cereal pathogens, including the powdery mildew pathogen Blumeria graminis and corn smut fungus Ustilago maydis, as well as the soybean rust fungus Phakopsora pachyrhizi (see poster). Multi-cellular infection cushions are mainly found in root pathogens, such as Rhizoctonia solani (see poster). In addition to fungi, filamentous oomycete pathogens, such as Phytophthora and Pythium species, form appressoria that are small compared to fungal appressoria, non-melanised and separated from their germ tubes by false septa. By contrast, in fungi, a septum normally separates the appressorium from the spore (Chethana et al., 2021a). Unlike a fungal appressorium, which enters the underlying plant cell at a 90° angle, penetration hyphae of the oomycete late blight pathogen Phytophthora infestans appear to enter at a diagonal angle, using a specific ‘naifu’ cutting action to breach the host surface. (Bronkhorst et al., 2021).

Relatively little is known about the evolutionary origin of appressoria. It has been proposed that they evolved first in mutualistic fungi (Mendgen et al., 1996); arbuscular mycorrhizal fungi, for example, develop similar hyphopodia to penetrate roots (Bonfante and Genre, 2010). Recent phylogenetic analysis suggests that simple non-pigmented appressoria appeared first, and later diversified into melanised or compound appressoria (Chethana et al., 2021b). However, there are open questions regarding the origin of proto-appressorium structures and their relationship to other differentiated non-hyphal structures, such as conidia, for example, or the terminally differentiated haustoria used by fungal pathogens to colonise plant cells after infection (Bozkurt and Kamoun, 2020), as well as resting bodies, such as sclerotia, which are often melanised (Mendgen et al., 1996).

Although our knowledge regarding the origin of appressoria is limited, recent breakthroughs have revealed key cellular mechanisms associated with their development. Here, we summarise these new insights with a special focus on the rice blast fungus M. oryzae, where appressoria have been intensively studied (Wilson and Talbot, 2009; Eseola et al., 2021).

Appressoria in the rice blast fungus are unicellular and differentiate from a polarised germ tube shortly after germination of an asexual spore called a conidium. In M. oryzae, conidia are three-celled tear-drop-shaped (pyriform) spores that stick to the hydrophobic waxy cuticle of rice leaves by secreting a glue called spore tip mucilage (Hamer et al., 1988). Once attached, the conidium germinates to produce a polarised germ tube that ceases apical growth within 4 h (Wilson and Talbot, 2009), flattening and hooking at its tip, which tightly adheres to the leaf surface. The fungus then secretes morphogenetic proteins called hydrophobins, such as Mpg1 and Mhp1 (Kim et al., 2005; Talbot et al., 1993), which polymerise, undergoing interfacial self-assembly, to form an amphipathic monolayer between germ tube and leaf surface, thereby enabling cutinases and mucilage to glue the developing cell to the cuticle (Kim et al., 2005; Pham et al., 2016; Talbot et al., 1993; Whiteford and Spanu, 2002). Hydrophobin self-assembly therefore acts as a primer for attachment factors – a prerequisite to efficient formation of an appressorium (Pham et al., 2016). Recent work has also highlighted the important role of spermine synthase (Ssp1) in acting as an antioxidant in the endoplasmic reticulum (ER), enabling production and secretion of spore tip mucilage necessary to appressorium adhesion (Rocha et al., 2020).

To develop an appressorium, the blast fungus perceives physical cues, such as surface hardness and hydrophobicity, and chemical signals, including leaf waxes and cutin monomers. G-protein coupled receptors (GPCRs) mediate recognition of these signals (Liu and Dean, 1997). Pth11, for example, is a GPCR involved in surface perception, which, when activated, leads to dissociation of the heterotrimeric Gα proteins MagA and MagB. When surface hydrophobicity is detected, the Gβγ subunit Mgb1 separates from the complex and binds to membrane-bound adenylate cyclase (Mac1), which catalyses generation of cyclic AMP (cAMP) (Adachi and Hamer, 1998). Rgs1 meanwhile regulates G-protein signalling and is required for correct perception of inductive surfaces (Liu et al., 2007; Ramanujam et al., 2012). cAMP-dependent protein kinase A signalling then regulates appressorium morphogenesis and turgor generation. In parallel, the sensor proteins Msb2 and Sho1 activate the Pmk1 mitogen-activated protein kinase (MAPK) pathway in response to surface hydrophobicity, cutin monomers and leaf waxes (Liu et al., 2011). Pmk1 pathway activation also involves Mgb1, as well as Ras2 and Cdc42 (Wilson and Talbot, 2009) (see poster). Mgb1 therefore plays important roles in both the cAMP and Pmk1 MAPK cascades, and likely regulates multiple steps in the morphogenetic transition required for infection (Nishimura et al., 2003). Additionally, Mgb1 interacts directly with Mst50, a putative scaffold protein that integrates multiple upstream signals to activate Pmk1 (Park et al., 2006), through its interaction with Gef1, Cdc42, Scd1, Ral2 and the Ras GAP protein Smo1 (Li et al., 2017; Qu et al., 2021). Activation of Pmk1 also requires the Mst7 MAPK kinase (MAPKK) and Mst11 MAPK kinase kinase (MAPKKK) (Zhao et al., 2005), which both bind to Mst50 (Park et al., 2006). The importance of Pmk1 to appressorium development is illustrated by the fact that Δpmk1 null-mutants, or those in which the kinase can be chemically inactivated, such as analogue-sensitive pmk1 mutants (pmk1AS) (Sakulkoo et al., 2018), fail to form appressoria or to infect rice plants, even when the surface cuticle is removed (Sakulkoo et al., 2018; Xu and Hamer, 1996). Furthermore, homologues of Pmk1 are necessary for plant infection in more than 20 fungal pathogens, including both appressorium- and non-appressorium-forming species (Jiang et al., 2018; Turrà et al., 2014). Pmk1 therefore likely serves a conserved regulatory function governing invasive fungal growth (Jiang et al., 2018; Turrà et al., 2014). Phosphorylation targets of Pmk1, including the transcription factors Mst12 and Hox7, have recently been identified (Osés-Ruiz et al., 2021). Hox7 is essential for appressorium formation, whereas Mst12 is required for penetration and invasive growth (Cao et al., 2016; Kim et al., 2009; Osés-Ruiz et al., 2021; Park et al., 2002; Yue et al., 2016). Pmk1 ultimately controls expression of more than 6500 genes that are differentially regulated during appressorium development (Osés-Ruiz et al., 2021), highlighting the morphogenetic complexity of appressorium formation.

Appressorium development is also tightly coordinated with cell cycle progression (see poster). Inhibition of S-phase impairs hooking and initiation of appressorium formation, while the G2-M transition is necessary for appressorium maturation (Osés-Ruiz et al., 2017; Saunders et al., 2010; Veneault-Fourrey et al., 2006). In B. graminis, inhibition of S-phase prevents maturation of appressoria (Hansjakob et al., 2012), while in U. maydis, arrest of filaments at the G2-phase checkpoint is necessary for appressorium formation (Castanheira et al., 2014).

Appressorium turgor in M. oryzae is generated by accumulation of glycerol and other polyols to high concentrations, which draws water into the cell by osmosis, creating turgor of up to 8.0 MPa (de Jong et al., 1997). The melanin-rich cell wall is impermeable to glycerol, but freely permeable to water which rapidly enters the cell. Compatible solutes used by other appressorium-forming pathogens have yet to be identified, but glycerol and mannitol are abundant in many fungi, and proline may be important in oomycete appressoria (Talbot, 2019). In M. oryzae, trehalose and glycogen are rapidly degraded during spore germination, and trehalose-6-phosphate synthesis is also required for fungal virulence (Foster et al., 2003), regulating an NADPH-dependent genetic switch that controls glucose-6-phosphate metabolism and nitrogen source utilisation (Wilson et al., 2010). This is vital for the metabolic changes necessary for invasion of plant tissue (Fernandez and Wilson, 2012; Wilson et al., 2010). During appressorium maturation, lipid bodies move to the developing appressorium in a Pmk1-dependent manner, and their degradation, and subsequent fatty acid metabolism and glycerol synthesis, requires the cAMP/PKA pathway (Wang et al., 2007). Mutants unable to synthesise melanin cannot retain glycerol, and do not generate turgor pressure or cause disease (Chumley and Valent, 1990). Most fungal melanins are derived from the precursor molecule 1,8-dihydroxynaphthalene (DHN), ultimately derived from acetate, which enters a series of sequential enzymatic steps via the enzymes Alb1, 4Hnr, Rsy1 and Buf1 (see poster). DHN monomers polymerise to yield melanin through the oxidative activity of laccases, and it is deposited in a layer within the appressorium cell wall (see poster). Melanin is critical for appressorium function in many species (Lin et al., 2012), but there are exceptions, such as the soybean rust fungus P. pachyrhizi, which produces hyaline non-melanised appressoria that can generate 5.13 MPa of turgor (Chang et al., 2014; Loehrer et al., 2014). Fungal appressoria have therefore evolved distinct means of pressure generation (Chang et al., 2014; Loehrer et al., 2014; Ludwig et al., 2014).

The final prerequisite for appressorium formation in M. oryzae is autophagic degradation of the contents of the three-cell conidium (see Box 1 and poster). Autophagy is normally considered a recycling, cell survival strategy (Klionsky, 2007), but it can also regulate a form of programmed cell death (Tsukada and Ohsumi, 1993). In M. oryzae, autophagy in the conidium is necessary for appressorium function (Kershaw and Talbot, 2009; Liu et al., 2016; Lv et al., 2017; Ying et al., 2016; Zhang et al., 2013; Zhou et al., 2020). Death of the conidium might, however, require an iron-dependent process called ferroptosis induced by accumulating lipid peroxides in the cell (Dixon et al., 2012; Kagan et al., 2017; Shen et al., 2020). Ferroptosis can be subverted by suppressing lipid peroxide levels within the cell using iron chelators to impair appressorium maturation (Liang et al., 2021; Shen et al., 2020; Shen and Naqvi, 2021).

The regulation of infection-associated autophagy also involves the target of rapamycin (TOR) signalling pathway (Beck and Hall, 1999; Conrad et al., 2014; Shashkova et al., 2015). In fungal pathogens, TOR influences the cAMP response and cell wall integrity pathway, as well as autophagy (López-Berges et al., 2010; Marroquin-Guzman and Wilson, 2015; Oh et al., 2008; Qian et al., 2018; Yu et al., 2014). In M. oryzae, the TOR pathway can be activated under high levels of intracellular glutamine and glucose, which inhibit appressorium formation. Inactive TOR is, therefore, thought to be necessary for cell cycle arrest and autophagy (Fernandez and Orth, 2018; Marroquin-Guzman et al., 2017; Marroquin-Guzman and Wilson, 2015). Furthermore, the serine/threonine phosphatase Tip41 mediates crosstalk between the TOR and the cell integrity pathway (Qian et al., 2018), mirroring a similar regulatory process in yeast (Helliwell et al., 1998; Kamada et al., 2005). Interestingly, in M. oryzae a VAD1 analogue of StAR-related lipid transfer (VASt-domain)-containing protein Vast1, monitors plasma membrane tension and sterol content of the plasma membrane, modulating the activity of TOR and thereby regulating autophagy (Zhu et al., 2021).

For successful penetration, it is necessary for the appressorium to switch from isotropic expansion to polarised growth. This requires reorganisation of the F-actin cytoskeleton at the appressorium pore, from which the penetration peg emerges (Bourett and Howard, 1992; Dagdas et al., 2012). F-actin forms a higher-order ring structure at the pore that requires septin guanosine triphosphatases (GTPases) (see poster). Septins are crucial for recruiting and organising F-actin, forming a hetero-oligomeric ring at the pore (Dagdas et al., 2012). Septins rigidify the cortex of the appressorium, scaffold actin and act as a lateral diffusion barrier for polarity determinants, such as Las17 (Dagdas et al., 2012; Dulal et al., 2020; Van Ngo and Mostowy, 2019). Likewise, the localisation of Rvs167, a BAR-domain-containing protein, is septin-dependent (Dagdas et al., 2012) and might enable the generation of membrane curvature that is associated with polymerisation of cortical F-actin (Mattila et al., 2007). Interestingly, organisation of the septin ring also relies on an intact microtubule cytoskeleton (Dulal et al., 2021). Septin interactions with phosphatidylinositol (PtdIns) 4-phosphate (PI4P) or phosphatidylinositol-4,5-bisphosphate (PI4,5P2) and the ezrin, radixin and moesin-like protein Tea1 at the plasma membrane also occur at the appressorium pore (see poster). Septin organisation requires the very-long-chain fatty-acid-rich (VLCFA) domains of PIPs, and inhibition of VLCFA biosynthesis prevents septin ring formation and plant infection, which has been used as a means of selecting broad-spectrum fungicides (He et al., 2020). This highlights the conserved nature of septin-dependent processes in the appressorium, because these fungicides control diverse plant and insect-pathogenic fungi (He et al., 2020).

Determining how the appressorium perceives a signal to repolarise is a long-standing question. In M. oryzae, the histidine-aspartate kinase Sln1 enables the appressorium to sense when a threshold of turgor has been reached so it can switch to polarised growth (Ryder et al., 2019). Sln1 was first identified as an osmosensor in yeast and regulates hyperosmotic adaptation through the Hog1 MAPK signalling pathway. The equivalent MAPK in M. oryzae, Osm1 is, however, not involved in appressorium turgor generation (Dixon et al., 1999) and plays a very different role during plant infection (Liu et al., 2020), so M. oryzae Sln1 must operate in a distinct manner to control appressorium turgor. Sln1 negatively regulates melanin biosynthesis and the cAMP-response pathway, and works in parallel with the protein kinase C1 (Pkc1)-dependent cell integrity pathway by phosphorylating the phosphodiesterase enzyme PdeH to modulate cAMP levels, which in turn controls lipolysis and glycerol production (Ryder et al., 2019). Pkc1 might also target the regulatory subunit of NADPH oxidase, a homologue of mammalian p67^phox (also known as NCF2), which is necessary for regulated synthesis of reactive oxygen species (ROS) required for septin-mediated cytoskeletal reorganisation (Ryder et al., 2019, 2013) (see poster). This observation is consistent with studies in humans, where it was demonstrated that PKC is required for phosphorylation of gp91^phox (also known as CYBB) and its binding to Rac2, p67^phox (NoxR in fungi) and p47^phox (also known as NCF1; Bem1 in fungi) (Raad et al., 2008).

Plant infection proceeds by emergence of a rigid penetration peg from the base of the appressorium (Bourett and Howard, 1990). Protrusive force generated by appressoria from Colletotrichum has been measured at 17 µN using an optical waveguide (Bechinger et al., 1999; Talbot, 2003), which is (easily) sufficient to rupture the cuticle and plant cell wall. During M. oryzae infection, the resulting primary invasive hypha does not, however, rupture the plasma membrane of the invaded epidermal cell, but instead the plant membrane is invaginated around the fungal hypha, forming a special compartment called the extra-invasive hyphal membrane (EIHM) (Kankanala et al., 2007; Yi and Valent, 2013). This is analogous to the extrahaustorial membrane (Bozkurt and Kamoun, 2020) and, in the same way, might provide a means of sequestering nutrients from plant cells while also evading recognition. A membrane-rich cap is visible at the tip of the penetration hypha and maintained throughout cell colonisation. This structure is termed the biotrophic interfacial complex (BIC) and appears to be a site for delivery of fungal effectors (Valent and Khang, 2010) (see poster). Effector proteins are secreted fungal proteins that target components of the plant immune system to suppress plant defence (Kamoun, 2006; Valent and Khang, 2010). M. oryzae effectors perturb a range of molecular functions, including suppressing chitin-triggered immunity (Mentlak et al., 2012) and reprogramming host transcription (Kim et al., 2020; Mentlak et al., 2011, 2012; Mosquera et al., 2009; Valent and Khang, 2010). Effectors destined for delivery into rice cells accumulate at the BIC, whereas those targeting extracellular immune responses are secreted from the hyphal tip into the apoplastic space, between the fungal cell wall and EIHM (Valent and Khang, 2010; Martin-Urdiroz et al., 2016). In contrast to apoplastic effectors, BIC-localised cytoplasmic effectors are secreted using an unusual Golgi-independent secretion system (Giraldo et al., 2013). Focal secretion of effectors directly from the base of appressoria has also been observed in Colletotrichum higginsianum (Kleemann et al., 2012) and C. orbiculare, where a membrane-rich structure similar to the BIC has been reported (Irieda et al., 2014; Khang et al., 2010; Valent and Khang, 2010).

When the fungus encounters the junction with a neighbouring cell, the hyphal tip undergoes cortical scanning until it locates a pit field, where plasmodesmata are located. The hyphal tip swells into a structure resembling an appressorium, called a transpressorium (Cruz-Mireles et al., 2021). The transpressorium undergoes hyphal constriction and forms a narrow infection peg that enters the adjacent cell. Remarkably, plant plasma membrane integrity is maintained during this process, with the EIHM surrounding the fungus in the newly invaded cell, where a BIC rapidly forms. There are clear parallels between appressoria and transpressoria, because Pmk1 regulates morphogenesis of both cell types, and transpressorium function is also mediated by septin proteins (Sakulkoo et al., 2018) (see poster).

Although the regulatory framework governing appressorium development in M. oryzae is known, there are very significant gaps in our understanding (listed on the poster). The precise mechanisms by which surface signals are perceived, for example, are far from clear. There may be a role for thigmotropic sensors and receptors of plant-derived compounds in addition to Pth11, for instance, and the ligand recognised by the Pth11 GPCR is so far unknown. How the upstream signalling pathways activate the Pmk1 pathway is also not at all clear and will require high-resolution proteomic methods, coupled with functional gene studies, to be fully deciphered. Similarly, downstream of Pmk1, we have little understanding of the transcriptional regulatory hierarchy that controls the major changes in gene expression required for appressorium morphogenesis. The immediate targets of Pmk1 are now being defined (Osés-Ruiz et al., 2021), but iterative rounds of phosphoproteomics and chromatin immunoprecipitation (ChIP)-seq analysis will be necessary to identify how the regulatory pathway actually operates. The precise role of TOR signalling also requires further attention, as it might impact on the control of plasma membrane dynamics and turgor control, as well as autophagy, during appressorium development. Understanding Sln1-dependent turgor sensing is also a priority, because it is pivotal to the control of cellular reorganisation and appressorium repolarisation. The direct targets of the Sln1 kinase have not yet been identified, nor its interplay with TOR, protein kinase C, Pmk1 and cell cycle regulation. Finally, the identification of the transpressorium generates many questions (see poster) regarding the parallels between both infection cells. M. oryzae is a valuable model for investigating the cell biology of appressoria, but its utility will only be fully realised if the conservation of many of the processes identified is confirmed in other appressorium-forming fungi. This has proceeded in some cases – Pmk1 being a clear example – but there is a need to investigate some of the experimentally recalcitrant appressorium-forming fungi, such as rusts (which are unculturable), to determine the extent to which the discoveries in rice blast can be more generally applied.